On the stereo- and regioselectivity of molybdenum-catalyzed allylic

On the Stereo- and Regioselectivity of. Molybdenum-Catalyzed Allylic Alkylations. Stereocontrolled Approach to Quaternary Carbons and. Tandem Alkylati...
1 downloads 0 Views 269KB Size
3343

J. Am. Chem. SOC. 1983, 105, 3343-3344

On the Stereo- and Regioselectivity of Molybdenum-Catalyzed Allylic Alkylations. Stereocontrolled Approach to Quaternary Carbons and Tandem Alkylation-Cycloaddition

Scheme I 0

Barry M. Trost* and Mark Lautens McElvain Laboratories of Organic Chemistry Department of Chemistry, University of Wisconsin Madison, Wisconsin 53706 Received December 29, 1982 Regioselectivity in allylic alkylations of stabilized carbon nucleophiles mediated by transition metals normally is dominated by steric approach leading to attack at the less substituted carbon of a ?r-allyl intermediate.' We report that in molybdenumcatalyzed allylic alkylation^^-^ the regiochemistry depends upon the nature of the nucleophile to an unusual degree. This unusual dependence of regioselectivity was first noted with 2-acetoxy-1-methylenecyclohexane (1) as shown in eq 1. Under

75%

E (E NOH or ESA' Molt1

1

otherwise identical conditions [NaH, 5% Mo(CO), (Mo(c)), PhCH3, 110 "C], the sodium salt of 2-carbomethoxycyclopentanone gave 2 while the sodium salt of an acyclic @-ketoester 3 and of dimethyl malonate 4 gave the products 54*5and 6,4 respectively, of exclusive attack at the ring carbon rather than the side-chain carbon. To explore the primarysecondary competition without complications of a ring system, the acyclic substrate 7 (Scheme I) was examined. Exactly the same trend was observed, although in this case a minor product identified as the alternative regioisomer was detected in the malonate case. This example also highlights the chemoselectivity by the reordering of the relative reactivity of the bromide and allylic acetate since in the absence of the molybdenum catalyst and in D M F solvent only bromide displacement is observed.' To explore this apparently contrathermodynamic result more fully, we turned our attention to the 4-tert-butyl- 1-vinyl- 1-acetoxycyclohexane case (8 Scheme 11). Even though the competition between the two termini of the allyl unit involves a primary vs. tertiary center, the anion of either dimethyl or di-tert-butyl malonate attacks the tertiary center exclusively to give lo4,*contaminated with a small amount of its stereoisomer (ratio 6:l). On the other hand, the corresponding 1,3-diketone l l a and @-keto ester l l b , and the C-methylated analogues of dimethyl malonate (llc) and the 1,3-diketone l l d led to exclusive terminal attack to give 12.4 (1) For reviews see: Trost, B. M.; Verhoeven, T. R. Compr. Organomet. Chem. 1982,8, 779-938. Trost, B. M. Pure Appl. Chem. 1981,53, 2357; Aldrichimica Acta 1981, 14, 43; Acc. Chem. Res. 1980, 13, 385. (2) Adams, R. D.; Chodosh, D. F.; Rosan, A. M.; Faller, J. W. J. Am. Chem. SOC.1979, 101, 2570. Faller, J. W.; Rosan, A. M . Ann. N.Y.Acad. Sci. 1977,295, 186. Also see: Bailey, N. A.; McCleverty, J. S . J. Chem. SOC., Chem. Commun. 1974, 592. (3) (a) Trost, B. M.;Lautens, M. J . Am. Chem. SOC.1982,104, 5543. (b) In this paper, we reported the Mo(bpy)-catalyzed alkylation of 3-acetoxy-4carboxaldehydocyclohex-1-ene.Subsequent control experiments reveal the reaction proceeds even in the absence of the Mo(bpy) catalyst. (4) This compound has been fully characterized spectrally and the elemental composition determined from high-resolution mass spectroscopy and/or combustion analysis. (5) The product is an equilibrium mixture of diastereomers, which were not separated. ( 6 ) BSA = 0,N-bis(trimethylsilyl)acetamide, E = C02CH3. (7) Cf.: Trost, B. M.; Verhoeven, T. R. J. Am. Chem. SOC.1980, 102, 4730. (8) The stereochemistry derives by comparison of the 'Hand "C NMR data to that of the ethyl ester series. See: House, H. 0.;Lubinkowski, J.; Good,J. J. J . Org. Chem. 1975, 40, 86.

5

+

Br(CH217CH=CHCHZCHE2 1

To demonstrate that the regioselectivity represented the intrinsic bias of the nucleophile and not subtle changes in catalyst, a competition between dimethyl malonate and dimethyl methylmalonate for limited amounts of 8 was conducted. Absolutely no crossover products were observed, i.e., malonate anion attacked only the tertiary position to give 10 and methylmalonate anion attacked only the primary carbon to give 12c! Nevertheless, a different catalyst can be generated when dimethyl malonate is employed. The allylic acetate 13 shows no reaction at all with dimethyl malonate but alkylates with dimethyl methylmalonate, albeit slowly (eq 2).4 The steroid substrate 14 NR

z

BrCH2l7CHCH=Cb I CHE2

CH2E2 SSAS 70%

&, -c 3"rL'(?) NaH + Mo~)

'2

73%

showed the same remarkable sensitivity of reactivity towards nucleophile. The source of the failure of alkylation with malonate anion and 13 or 14 appears to be due to the generation of a

4 fl ~~

NR

(: + NaHor BSAS MOiCI

___)

NOH Mot!

'5

50%

13'

different catalyst derived from the reaction of Mo(c) and dimethyl malonate such as (C0)6-nMo[CHE2]: or (CO)5Mo[CHE2]-.10 With the slow to react substrates 13 and 14, such catalysts appear to be of insufficient reactivity to initiate ionization. However, with the more reactive substrates, reaction proceeds normally. The source of the unusual divergence in regioselectivity between dimethyl malonate and other nucleophiles does not appear to stem from a different catalyst as shown by the competition experiment. It would appear to stem from an intricate balance among reactivity of the anion, its steric demands, the charge distribution in the intermediate wallyl system, and the stability of the resultant olefin as well as the olefin-metal complex. In an unsymmetrical complex such as 16, C(a) would be more electron deficient than C(b). Nu

Further, attack at C(a) leads to the sterically and electronically preferred olefin-Mo complex as the initial product. The higher reactivity of the malonate anion and its minimum steric requirements lead these two factors to dominate, and path 3a is traversed. On the other hand, introduction of a single alkyl substituent makes the steric demands of the nucleophile sufficiently important that path 3b totally dominates. Similarly, reducing the reactivity of the anion as in the case of a 1,3-diketone di(9) Cf.: Dunne, T. G.; Cotton, F. A. Inorg. Chem. 1963, 2, 2633. (10) Cf: Darensbourg, M. Y.; Conder, H. L.;Darensbourg, D. J.; Hasday, C. J. Am. Chem. SOC.1973,95,5919. Darensbourg, D. J.; Darensbourg, M. Y. Inorg. Chem. 1970,9, 1691. For pioneering work in this general field see: Fischer, E. 0.; Maasbol, A. Angew. Chem., Int. Ed. Engl. 1964, 3, 580. Fischer, E. 0.;Kiener, V. J. Organomet. Chem. 1970, 23, 215.

0002-7863/83/1505-3343$01,50/00 1983 American Chemical Society

3344

J . Am. Chem. SOC.1983, 105, 3344-3345

Scheme I1

OR

dR2 '2

Scheme I11

templates provide a most powerful tool for allylic alkylations.

Acknowledgment. We thank the National Science Foundation for their generous support of our programs, the NSERC of Canada for a graduate fellowship for M.L., and Brian Petersen for preparation of several starting materials. We thank Pressure Chemical Co. and Climax Molybdenum Co. of Michigan for gifts of molybdenum hexacarbonyl. Registry NO.1, 53723-50-5; 2, 82352-45-2; 3, 85390-61-0; 4, 18424-

8546

83%

75%

minishes the importance of the charge distribution in 16 and leads to preferential formation of the thermodynamically more stable olefin. Finally, a 8-keto ester, intermediate in reactivity, shows intermediate behavior. When steric factors dominate as with the anion of 2-carbomethoxycyclopentanone, only path 3b is observed. On the other hand, the keto ester l l b prefers attack at the more hindered position (Le., formation of 5) unless the steric requirements of the allyl system overwhelms as in the case of allyl acetate 8. The complementary regiochemical behavior of malonate and substituted malonates offer interesting applications. As shown in the case of 8 and the steroid substrate 17, a high preference $8'17

76-5; 5, 85390-62-1; 6,67428-13-1; 7, 85390-63-2; 8, 85390-64-3; 9, 85390-65-4; 10,85390-66-5; 110, 123-54-6; llb, 30414-55-2; llc, 60902-9; lld, 815-57-6; 120,85390-67-6; 12b,85405-71-6; 12q 85390-68-7; 12d, 85390-69-8; 13,85390-70-1; 14, 85390-71-2; 15,85390-72-3; 17, 85390-73-4; 18,7014-63-3; Mo(CO),, 13939-06-5; methyl 2-hydroxyI-cyclopentene-1-carboxylate-Na, 63 178-03-0; methyl l-(lO-bromodec2-enyl)-2-oxo-l-cyclopentanecarboxylate,85390-74-5; methyl 10bromo-3-ethenyl-2-methoxycarbonyldecanoate, 85390-75-6; methyl 12bromo-2-methoxycarbonyidodec-4-enoate,85390-76-7; methyl 2-oxocyclopentanecarbaxylate, 10472-24-9; methyl 5-(cyclohex-3-en-l-yI)2,4-dimethyl-2-methoxycarbonylpe.nt-4-enoate, 85390-77-8; 3a-ethenyl38- [bis(methoxycarbon yl)methyl] - Sa-cholestane, 8 5390-7 8-9; methyl 8-ethenyl-a-methoxycarbonyl1-cyclohexene-1-propanoate, 85390-79-0; methyl 5-(cyclohex-l-en-l-yl)-2-methyl-2-methoxycarbonyl-4-pentenoate, 85390-80-3; 7-(cyclohex- 1-en- 1-yl)-4,4-bis(methoxycarbonyl)1,6-heptadiene, 85390-81-4; 2,2-bis(methoxycarbonyl)2,3,3a,6,7,8,9,9a,10,lOa-decahydro-l~-benz~indene, 85390-82-5.

Lewis Acid Catalysis of Coumarin Photodimerization Frederick D. Lewis,* Daniel K. Howard, and Joe D. Oxman

for equatorial attack of malonate creates a quaternary center with high stereoselectivity' and independent of the stereochemistry of the starting allyl acetate13-aresult that permits a single product to result from a stereoisomeric mixture of starting materials. The failure of an additional double bond as in diallyl acetate 18, to change the established trends permits a tandem alkylation-cycloaddition as shown in Scheme IIL4 The fact that 18 arises so simply by addition of a vinyl organometallic to an a,@unsaturated carbonyl system makes this an attractive strategy for making substrates capable of an intramolecular Diels-Alder reaction.14 These results demonstrate that by appropriately choosing the nucleophile, great regiochemical control can be exercised in Mo(c)-catalyzed reactions. Combined with the earlier observation3 that the choice of ligands also affects regiochemistry, molybdenum (1 1) For an excellent review see: Martin, S. F.; Tetrahedron 1980,36,419. For a case using Grignard reagents and allyl alcohols with nickel catalysts see: Buckwalter, B. L.; Burfitt, I. R.; Felkin, H.; Joly-Gondket, M.; Naemura, K.; Salomon, M. F.; Wenkert, E.; Wovkulich, P.M. J. Am. Chem. Soc. 1978,100, 6445. (12) The stereochemistry of the alkylation product of 17 is assigned by analogy to 10. (13) As in the case of 8, 10 is by far the major product, but the alternative stereoisomer is also observed. (14) Cf.: Brieger, G. J. Am. Chem. SOC.1963, 85, 3783. Corey, E. J.; Glass, R. S . J. Am. Chem. SOC.1967, 89, 260. Frater, G. Helv. Chim. Acta 1974, 57, 172. Bajorek, J. J. S.; Sutherland, J. K. J . Chem. SOC.,Perkin Trans. 1 1975,1559. Oppolmr, W.; Snowden, R. L. Tetrahedron Lett. 1976, 4187.

Department of Chemistry, Northwestern University Evanston, Illinois 60201 Received November 12, I982 Lewis acids have been widely employed as catalysts for thermal Diels-Alder and ene reactions, especially those involving a,8unsaturated esters.lJ The e n h a n d reactivity and stereoselectivity observed in many such reactions have been attributed to changes in frontier orbital energies and C = C double bond polarity upon complexation of the carbonyl oxygen.2 We recently reported that the spectroscopic properties and unimolecular photoisomerization reactions of several a,@-unsaturatedesters are profoundly changed by complexation with Lewis acids such as BF3, EtAlCl,, and SIIC~,.~ The possibility that Lewis acids might also serve as catalysts for photochemical [2 + 21 cycloaddition reactions was suggested many years ago by reports concerning the photodimerization of dibenzylideneacetone in the presence of U02+and S ~ I C and ~ , ~is borne out by investigations currently in progress in our laboratory. We report here on the pronounced effects of the Lewis acid BF3 on the efficiency and regiochemistry of coumarin photodimerization. The photodimerization of coumarin has been the subject of numerous preparative and mechanistic investigations.5 The (1) Flemming, I. "Frontier Orbitals and Organic Chemical Reactions"; Wiley: New York, 1976,; pp 214-223. (2) Houk, K. N.; Strozier,R. W. J . Am. Chem. Soc. 1973,95,4094-4096. (3) Lewis, F. D.; Oxman, J. D. J. Am. Chem. Soc. 1981,103,7345-7347. (4) (a) Stobbe, H.; Farber, E. Chem. Eer. 1925,58, 1548-1553. (b) Alcock, N. W.; Herron, N.; Kemp. T. J.; Shoppee, C. W. J. Chem. SOC., Chem. Commun. 1975, 785-186.

0002-786318311505-3344$01.50/00 1983 American Chemical Society